Method for processing multi-source data

Abstract

A method for processing multi-source data includes: generating a private key fragment of the data providing node, as well as an encryption public key and an evaluation public key, in which, other participating nodes include other data providing nodes and a data using node; generating ciphertext data by encrypting local sample data with the encryption public key, and sending the ciphertext data to a coordination node for training a machine learning model; receiving ciphertext model parameters corresponding to model parameters of the machine learning model trained by the coordination node, in which, the ciphertext model parameters are determined by the coordination node based on the evaluation public key and the ciphertext data received; and decrypting the ciphertext model parameters with the private key fragment, and sending a decrypted model parameter fragment to the data using node.

What is claimed is: 
     
       1. A method for processing multi-source data, applicable for a data providing node, comprising:
 generating a private key fragment of the data providing node, an encryption public key and an evaluation public key, wherein, other participating nodes comprise other data providing nodes and a data using node; 
 generating ciphertext data by encrypting local sample data with the encryption public key, and sending the ciphertext data to a coordination node for training a machine learning model; 
 receiving ciphertext model parameters corresponding to model parameters of the machine learning model trained by the coordination node, wherein, the ciphertext model parameters are determined by the coordination node based on the evaluation public key and the ciphertext data received; and 
 decrypting the ciphertext model parameters with the private key fragment, and sending a decrypted model parameter fragment to the data using node. 
 
     
     
       2. The method according to  claim 1 , wherein generating the private key fragment of the data providing node, the encryption public key and the evaluation public key comprises:
 generating the private key fragment, and generating an encryption public key fragment of the data providing node and a target intermediate result corresponding to the evaluation public key based on the private key fragment; 
 broadcasting the encryption public key fragment and the target intermediate result; 
 receiving respective encryption public key fragments and target intermediate results sent by other participating nodes; and 
 generating the encryption public key based on respective encryption public key fragments, and generating the evaluation public key based on respective target intermediate results. 
 
     
     
       3. The method according to  claim 2 , after generating the evaluation public key, further comprising:
 sending the evaluation public key to the coordination node. 
 
     
     
       4. The method according to  claim 2 , wherein generating the private key fragment of the data providing node comprises:
 acquiring key generation parameters, wherein, the key generation parameters comprise a public parameter and a public random number; and 
 acquiring the private key fragment based on the public parameter and a private key generation algorithm. 
 
     
     
       5. The method according to  claim 4 , wherein generating the encryption public key fragment of the data providing node and the target intermediate result corresponding to the evaluation public key based on the private key fragment comprises:
 generating the encryption public key fragment based on the private key fragment, the public random number and an encryption public key generation algorithm; and 
 generating the target intermediate result based on the private key fragment and an evaluation public key generation algorithm. 
 
     
     
       6. The method according to  claim 5 , wherein generating the target intermediate result based on the private key fragment and the evaluation public key generation algorithm comprises:
 generating a first intermediate result of the evaluation public key based on the private key fragment, the public random number and the evaluation public key generation algorithm, and broadcasting the first intermediate result; 
 receiving respective first intermediate results sent by other data providing nodes; 
 acquiring a second intermediate result of the evaluation public key based on the first intermediate result of the evaluation public key and the first intermediate results of other data providing nodes; and 
 acquiring the target intermediate result of the evaluation public key based on the private key fragment, the public random number and the second intermediate result, and broadcasting the target intermediate result. 
 
     
     
       7. An electronic device, comprising a processor and a memory; wherein the processor runs a program corresponding to an executable program code by reading the executable program code stored in the memory to implement the method of  claim 1 . 
     
     
       8. A computer-readable storage medium with a computer program stored thereon, wherein the program is configured to implement the method of  claim 1 . 
     
     
       9. A method for processing multi-source data, applicable for a data using node, comprising:
 generating a private key fragment of the data using node, an encryption public key and an evaluation public key, wherein, other participating nodes comprise data providing nodes; 
 generating ciphertext data by encrypting local sample data with the encryption public key, and sending the ciphertext data to a coordination node for training a machine learning model; 
 receiving ciphertext model parameters corresponding to model parameters of the machine learning model trained by the coordination node, wherein, the ciphertext model parameters are determined by the coordination node based on the evaluation public key and received ciphertext data of the data providing node; 
 acquiring a first model parameter fragment by decrypting the ciphertext model parameters with the private key fragment; 
 receiving second model parameter fragments sent by other participating nodes; and 
 acquiring the machine learning model based on the first model parameter fragment and the second model parameter fragments. 
 
     
     
       10. The method according to  claim 9 , wherein generating the private key fragment of the data using node, the encryption public key and the evaluation public key comprises:
 generating the private key fragment of the data using node, and generating an encryption public key fragment of the data using node based on the private key fragment; and 
 broadcasting the encryption public key fragment. 
 
     
     
       11. The method according to  claim 10 , wherein generating the private key fragment of the data using node comprises:
 acquiring key generation parameters, wherein, the key generation parameters comprise a public parameter and a public random number; and 
 acquiring the private key fragment based on the public parameter and a private key generation algorithm. 
 
     
     
       12. An electronic device, comprising a processor and a memory; wherein the processor runs a program corresponding to an executable program code by reading the executable program code stored in the memory to implement the method of  claim 9 . 
     
     
       13. A computer-readable storage medium with a computer program stored thereon, wherein the program is configured to implement the method of  claim 9 .